HOW SCIENCE DEALS WITH POSSIBILITY RATHER THAN A CLEARCUT CAUSE AND EFFECT
A few years ago I wrote a Life Lines column about Napoleon’s death. For several decades historians believed Napoleon was poisoned by the addition of arsenic to his food while he was in exile at St. Helena. As evidence they cited studies of a sample of his hair, bombarded with neutrons, which demonstrated a high incidence of arsenic in the hairs. I felt at that time, that was not a sufficient proof and I cited a news article I had remembered reading a generation earlier when Clare Booth Luce was an Ambassador to Italy and lived in Rome. She got quite sick from what turned out to be arsenic poisoning caused by her inhaling arsenic dust that flaked off from the paint in her room.
In a recent issue of Science, I noted that an Italian study of hair from Napoleon as a child as well as hair from members of his family and other contemporaries who never went to St. Helena, showed a similar high amount of arsenic in their hairs. Arsenic was widely used in paints, cosmetics, and as food additives in those days. The Italian Institute of Nuclear Physics concluded that homicide could be ruled out in the face of this widespread arsenic contamination of French middle class and wealthy people. To prove homicide something other than the presence of arsenic in the hair would be required.
The presence of arsenic in the hair is at best a necessary but not sufficient evidence of the deliberate poisoning of Napoleon. Most physicians who examined Napoleon’s body after he died concluded he died of stomach cancer although in those days there was no cell theory to back that up. While we may never know what caused Napoleon’s death, we at least can cast doubt that he was murdered. We crave certitude but science tells us we are limited to what we know and can’t let speculation substitute for knowledge. This is why the problem of Agent Orange exposure (or its dioxin contaminant) is so difficult. The presence of dioxins in veterans at parts per trillion (which is a very small amount compared to Napoleon’s arsenic) is a necessary but not sufficient piece of evidence that a veteran’s cancer, or enlarged prostrate, or diabetes is associated with that exposure. Because there is no proof that Agent Orange exposure causes human health problems other than chloracne (a skin disease) the National Academy of Sciences in the US uses a statistical association. If there are more veterans than predicted with a particular type of cancer, then those veterans are given the benefit of the doubt and their medical costs are covered. But if a veteran exposed to Agent Orange gets, let us say bladder cancer, and there is no excess association between those veterans and bladder cancer, that veteran does not get government provided medical care. The problem is really more social than scientific. If all people (or veterans) were covered by a national health insurance, then all people would have government-provided reimbursement for health care. We have done that with Social Security for retirement since the 1930s and someday we will extend that outlook to our health. Proof requires standards of science that are sometimes difficult to achieve, like the effects of low doses of radiation on cancer or mutation production, the effect of microwaves on cancer or mutation induction, and the possible health hazards of hundreds of chemicals in our environments and diets. In general harm from heavy, chronic or acute doses of exposure are easier to prove or disprove.